“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

Deep learning for healthcare applications based on physiological signals: A review.

Hand movement classification based on surface electromyography (sEMG) pattern recognition is a promising approach for upper limb neuroprosthetic control. However, maintaining day-to-day performance is challenged by the non-stationary nature of sEMG in real-life operation. In this study, we propose a self-recalibrating classifier that can be automatically updated to maintain a stable performance over time without the need for user retraining. Our classifier is based on convolutional neural network (CNN) using short latency dimension-reduced sEMG spectrograms as inputs. The pretrained classifier is recalibrated routinely using a corrected version of the prediction results from recent testing sessions. Our proposed system was evaluated with the NinaPro database comprising of hand movement data of 40 intact and 11 amputee subjects. Our system was able to achieve ~10.18% (intact, 50 movement types) and ~2.99% (amputee, 10 movement types) increase in classification accuracy averaged over five testing sessions with respect to the unrecalibrated classifier. When compared with a support vector machine (SVM) classifier, our CNN-based system consistently showed higher absolute performance and larger improvement as well as more efficient training. These results suggest that the proposed system can be a useful tool to facilitate long-term adoption of prosthetics for amputees in real-life applications.

/pdf/self-recalibrating-surface-emg-pattern-recognition-for-4ckrfmrs26.pdf

Self-Recalibrating Surface EMG Pattern Recognition for Neuroprosthesis Control Based on Convolutional Neural Network.

The surface electromyography (sEMG)-based gesture recognition with deep learning approach plays an increasingly important role in human-computer interaction. Existing deep learning architectures are mainly based on Convolutional Neural Network (CNN) architecture which captures spatial information of electromyogram signal. Motivated by the sequential nature of electromyogram signal, we propose an attention-based hybrid CNN and RNN (CNN-RNN) architecture to better capture temporal properties of electromyogram signal for gesture recognition problem. Moreover, we present a new sEMG image representation method based on a traditional feature vector which enables deep learning architectures to extract implicit correlations between different channels for sparse multi-channel electromyogram signal. Extensive experiments on five sEMG benchmark databases show that the proposed method outperforms all reported state-of-the-art methods on both sparse multi-channel and high-density sEMG databases. To compare with the existing works, we set the window length to 200ms for NinaProDB1 and NinaProDB2, and 150ms for BioPatRec sub-database, CapgMyo sub-database, and csl-hdemg databases. The recognition accuracies of the aforementioned benchmark databases are 87.0%, 82.2%, 94.1%, 99.7% and 94.5%, which are 9.2%, 3.5%, 1.2%, 0.2% and 5.2% higher than the state-of-the-art performance, respectively.

/pdf/a-novel-attention-based-hybrid-cnn-rnn-architecture-for-semg-2htxnz5meh.pdf

A novel attention-based hybrid CNN-RNN architecture for sEMG-based gesture recognition.

Deep Learning (DL) has recently become a topic of study in different applications including healthcare, in which timely detection of anomalies on Electrocardiogram (ECG) can play a vital role in patient monitoring. This paper presents a comprehensive review study on the recent DL methods applied to the ECG signal for the classification purposes. This study considers various types of the DL methods such as Convolutional Neural Network (CNN), Deep Belief Network (DBN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). From the 75 studies reported within 2017 and 2018, CNN is dominantly observed as the suitable technique for feature extraction, seen in 52% of the studies. DL methods showed high accuracy in correct classification of Atrial Fibrillation (AF) (100%), Supraventricular Ectopic Beats (SVEB) (99.8%), and Ventricular Ectopic Beats (VEB) (99.7%) using the GRU/LSTM, CNN, and LSTM, respectively.

A Review on Deep Learning Methods for ECG Arrhythmia Classification

A high performance electrocardiogram (ECG)-based arrhythmic beats classification system is presented in this paper. The classifier was designed based on convolutional neural network (CNN). Single channel ECG signal was segmented into heartbeats in accordance with the changing heartbeat rate. The beats were transformed into dual beat coupling matrix as 2-D inputs to the CNN classifier, which captured both beat morphology and beat-to-beat correlation in ECG. A systematic training beat selection procedure was also proposed which automatically include the most representative beats into the training set to improve classification performance. The classification system was evaluated for the detection of supraventricular ectopic beats (SVEB or S beats) and VEB using the MIT-BIH arrhythmia database. Our proposed method has demonstrated superior performance than several state-of-the-art detectors. In particular, our proposed CNN system has improved sensitivity and positive predictive rate for S beats by more than 12.2% and 11.9%, respectively, over these top performing algorithms. Our proposed CNN classifier with an automatic training beats selection process has shown to outperform the previous methods. The classifier is also a personalized one by combining training set from a common pool and a subject-specific set of ECG data. Our proposed system provides a reliable and fully automatic tool for detection of arrhythmia heartbeat without the need for manual feature extraction or expert assistant. It can potentially be implemented on portable device for the long-term monitoring of cardiac arrhythmia.

Automated ECG Classification Using Dual Heartbeat Coupling Based on Convolutional Neural Network

In this paper, we propose a semi-supervised learning-based ECG classification system for detection of supraventricular ectopic beats (SVEB or S beats) and ventricular ectopic beats (VEB or V beats) which does not require manual labeling of the patient-specific ECG data. Owing to inter-subject variability in ECG signal, patient-specific data is usually required to achieve good performance in ECG classification system. However, manual labeling of patient-specific data requires expert intervention, which is costly and time consuming. Our proposed system is based on a 2D convolutional neural network (CNN) with inputs generated from heartbeat triplets. The system also consists of two auxiliary modules: a normal beat estimation module and an iterative beat label update algorithm. The normal beat estimation selects a small amount of patient-specific normal beats accurately from the testing ECG record in an unsupervised manner. These estimated normal beats are used, together with a common pool dataset, to train a preliminary patient-specific CNN classifier which provides initial labels for the testing data. These labels then undergo a semi-supervised iterative update process for improved performance. Our proposed system was evaluated on the MIT-BIH arrhythmia database. The training of our proposed system is fully automatic, and its performance is comparable with several state-of-art supervised methods which require extra manual labeling of patient-specific ECG data. Our proposed system can be a useful tool for batch processing a large amount of ECG data in clinical applications.

Semi-supervised learning for ECG classification without patient-specific labeled data

Hand gesture recognition from forearm surface electromyography (sEMG) is an active research field in the development of motor prosthesis. Studies have shown that classification accuracy and efficiency is highly dependent on the features extracted from the EMG. In this paper, we show that EMG spectrograms are a particularly effective feature for discriminating multiple classes of hand gesture when subjected to principal component analysis for dimensionality reduction. We tested our method on the Ninapro database which includes sEMG data (12 channels) of 40 subjects performing 50 different hand movements. Our results demonstrate improved classification accuracy (by ∼10%) over purely time domain features for 50 different hand movements, including small finger movements and different levels of force exertion. Our method has also reduced the error rate (by ∼12%) at the transition phase of gestures which could improve robustness of gesture recognition when continuous classification from sEMG is required.

Short latency hand movement classification based on surface EMG spectrogram with PCA

A generative adversarial network (GAN) based fully automatic electrocardiogram (ECG) arrhythmia classification system with high performance is presented in this paper. The generator (G) in our GAN is designed to generate various coupling matrix inputs conditioned on different arrhythmia classes for data augmentation. Our designed discriminator (D) is trained on both real and generated ECG coupling matrix inputs, and is extracted as an arrhythmia classifier upon completion of training for our GAN. After fine-tuning the D by including patient-specific normal beats estimated using an unsupervised algorithm, and generated abnormal beats by G that are usually rare to obtain, our fully automatic system showed superior overall classification performance for both supraventricular ectopic beats (SVEB or S beats) and ventricular ectopic beats (VEB or V beats) on the MIT-BIH arrhythmia database. It surpassed several state-of-art automatic classifiers and can perform on similar levels as some expert-assisted methods. In particular, the F1 score of SVEB has been improved by up to 10% over the top-performing automatic systems. Moreover, high sensitivity for both SVEB (87%) and VEB (93%) detection has been achieved, which is of great value for practical diagnosis. We, therefore, suggest our ACE-GAN ( G enerative A dversarial N etwork with A uxiliary C lassifier for E lectrocardiogram) based automatic system can be a promising and reliable tool for high throughput clinical screening practice, without any need of manual intervene or expert assisted labeling.

Xiaolong Zhai

Papers

Self-Recalibrating Surface EMG Pattern Recognition for Neuroprosthesis Control Based on Convolutional Neural Network.

Automated ECG Classification Using Dual Heartbeat Coupling Based on Convolutional Neural Network

Semi-supervised learning for ECG classification without patient-specific labeled data

Short latency hand movement classification based on surface EMG spectrogram with PCA

Fully automatic electrocardiogram classification system based on generative adversarial network with auxiliary classifier